I do not know in detail about Buddhists' meditation but it fits nicely with the fourth scenario of learning new thoughts, feelings and actions... — Damir Ibrisimovic
In all 4 scenarios, the boundary of the "I" is the total physical animal; its own awareness of itself, its internal modeling of itself, is secondary, and it doesn't matter if that happens some time after the brain machinery has worked to produce whatever action it produces.
The Buddhist (and also general "non-dual" - as it's called - Asian philosophical) idea is that we are accustomed to thinking of "I" as an internal receiver of impressions, perceptions, experiences, etc., conceived of either as a sort of notional point on which all experiences impinge (by analogy with artistic perspective - which not coincidentally was being developed in a big way in the West roundabout the same time that modern philosophy and science developed) or as a kind of aware, awake "capacity" or "space" (this is the metaphor that's more usually favoured in the East).
If you reflect on your everyday experience, it normally seems as if the true "I" is something behind the eyes, looking out at the world through them, sort of prisoner inside the head in some sense, or inside the body, along for the ride. In Christianity, something like this idea is meant by the "soul," the thing that's notionally free in terms of the classical free-will debates.
This is an illusion, and we all have it (or most of us do, most of the time). A large part of the purpose of Eastern meditational systems is to either knock the illusion out of commission (considered a lesser result, because it's usually temporary) or to get into a position where you may still have the illusion, but you don't take it seriously, you don't habitually live from it (considered the greater result, usually called "enlightenment").
The Libet experiments only pose a problem if you believe that the "I" is something like this "soul," this "ghost in the machine" If no such thing exists, then the various timings of what goes on within the skin bag are of little importance to the question of free will, the cash-value of which is really more in the libertarian or political sense: one's will is free when it is not coerced - that is to say, this human animal, and all the workings within its skin bag, is the active entity, the entity whose will can be free or not, and either subject to coercion or not (one may also consider freedom from various kinds of
internal incoherences too, which have an effect somewhat analogous to external coercion, in that they also limit one's potential to act).
To put all this another way, when one says "I" one may, on the one hand, be referring to one's sense of oneself as the "ghost," but on the other hand, one may be referring to oneself as the human animal that one is. In the former case, there is no free will, because the entity in question simply doesn't exist (to either have free will or not) and all the puzzles arising from that are nugatory. If the latter sense, then all
ordinary talk about free-will makes perfectly ordinary, intelligible sense, even taking into account findings like Libet's.